Do Audio Speakers Break-in?

[charleski]

If “burn-in” changes the parameters of a component, how come does it always do for the better?

This always confused me as well. When a manufacturer is selling an expensive device one would hope they perform some degree of quality control at the factory, and test that it is performing to spec before it leaves the factory. If they think that their product changes for the better as a result of ‘burn-in’ does this mean that their testing spec is deliberately degraded to account for this?

When you consider the 4- or 5-figure price attached to many of these products, you have to ask why the manufacturer isn’t performing any required ‘burn-in’ before shipping it out, so they can test it does actually perform as desired in its final state.

 

[valve_integrator]

An example of a speaker manufacturer showing "green" and burn in changes and detailing the process is here:

https://www.wavecor.com/html/wf182bd03_04_07_08.html


Thor

Made in PRC also means highly variable QC and a very approximate understanding of what constitutes "good sound".
If you have ever been there and experienced their concert hall PA systems or even their concert halls.....arrrghhh WTF!

says it all really "exiting".....had me paused for a minute, then realised they meant EXCITING.

WTF are they doing running 20hz in a driver an entirely arbitrary level - well below resonant frequency?
It's considered Chinese "useful" like their crappy valve amps, and their crappy "great wall" wine, or simply their "great wall" so badly maintained it's falling apart?
So much from PRC resembles some giant north african bazaar, - you don't even know if you get what you pay for....

After burn-in specifications are measured 12 hours after exiting the transducer by a 20 Hz sine wave for 2 hours at level 10/14.1 VRMS (4/8 ohm version).

 

[OnLyTNT]

I don't remember seeing anything in data sheets for "burn in" of electronic or electric components. Not resistors, not caps, not inductors, not transistors, not rectifiers, not op amps, not DAC ICs, not microprocessors, not memory ICs..

But maybe I just missed it.

In addition, all electronic components have a life cycle, for microprocessors it's so long that they hardly fail in normal conditions but still they will reach to end of their life at some point. It's not like we perceive as "burn in". We can not use that term I think, may be "material degradation" I don't know. Also, they don't get any better, it's not like "ohhh this resistor after burn in period performs better", nothing like that. Especially caps degrades quickly and fail brutally, I can't remember exactly but 10-15 years before cap quality was a big issue. I was periodically replacing caps that failed on my motherboard, graphics card, psu, monitor list goes on

So, to my understanding if burn in happens for a product in a short term, that indicates low build quality and the product will much more quickly fail in future. That's my point of view.

Fun fact by the way, all atoms radioactively decay, so whatever you built will fail eventually in millions or billions years

I mean, how can one tune a bass port if the speaker resonance point keeps on changing?

Exactly!

If “burn-in” changes the parameters of a component, how come does it always do for the better? I never heard anyone complaining that after 201 hours, 32 minutes and 52 seconds of burn-in a speaker was sounding bad.

Agreed with your question. if the characteristic of a product changes that dramatically, what is going to happen in long term? Will it get better and better or will it go down hill?

 

[restorer-john]

Made in PRC also means highly variable QC and a very approximate understanding of what constitutes "good sound".
If you have ever been there and experienced their concert hall PA systems or even their concert halls.....arrrghhh WTF!

says it all really "exiting".....had me paused for a minute, then realised they meant EXCITING.

WTF are they doing running 20hz in a driver an entirely arbitrary level - well below resonant frequency?
It's considered Chinese "useful" like their crappy valve amps, and their crappy "great wall" wine, or simply their "great wall" so badly maintained it's falling apart?
So much from PRC resembles some giant north african bazaar, - you don't even know if you get what you pay for....

After burn-in specifications are measured 12 hours after exiting the transducer by a 20 Hz sine wave for 2 hours at level 10/14.1 VRMS (4/8 ohm version).

It's pretty obvious they measure parameters and likely match (pass/fail) drivers after running them for 2 hours. Sounds like good practice to me.

 

[Klonatans]

I have bought plenty of brand new speakers like Canton, Dali, KEF, ELAC, Quadral, Adam Audio and never ever anything has audibly changed even a tiny bit after lengthy period of use. The same applies to headphones: if I didn't like the sound signature in the beginning I never liked it after the so called burn-in period either. I think the most ridiculous thing is the burn-in cable supplied with Ibasso DX80 digital audio player.

 

[Keith_W]

It's pretty obvious they measure parameters and likely match (pass/fail) drivers after running them for 2 hours. Sounds like good practice to me.

For the low price their electronics sell at, one wonders what the tolerances and how much QC they perform.

 

[Thorsten Loesch]

I don't remember seeing anything in data sheets for "burn in" of electronic or electric components.

No, it doesn't state "burn in". That is a colloquial term. You will however find a lot of interesting stuff in small print.

Meanwhile something to read:

Douglas Self "Self-improvement for capacitors: the linearisation of polyester capacitors" (Linear Audio, Volume 1, April 2011)

Thor

 

[Thorsten Loesch]

Made in PRC also means highly variable QC

I'm normally totally for dumping on China and it's "engineering" at every corner, but the same issues exist in speaker drivers not made in China, if such a thing actually exists.

As with all things, in China (if you do your own QA/QC) you get what you pay for. Wave or make a "premium price" (and very high quality) product.

Even cheap factories for something as primitive as speaker drivers deliver great consistency, if you let them do enough testing on the line and use good processes. Mind you, that does cost extra.

Thor

 

[Thorsten Loesch]

If you have ever been there and experienced their concert hall PA systems or even their concert halls.....arrrghhh WTF!

I have been at the mixing desk for concerts in china, with all Chinese made gear.

I was able to obtain excellent sound quality, no acoustic feedback, monitors and backline excellent clarity and way loud enough according to the musicians, excellent sound in the house.

The hardware is generally pretty decent, it's the wet ware that is at fault and reason for the usually absolutely terrible results.

WTF are they doing running 20hz in a driver an entirely arbitrary level - well below resonant frequency?

Precisely what they are doing. And the level is NOT arbitrary. The engineers at Wavecor are not Chinese.

So much from PRC resembles some giant north african bazaar, - you don't even know if you get what you pay for....

When I was in Africa their bazaars were full of the lowest end cheapest china made garbage. Maybe that is why.

In China you must make sure you get what you pay for, or they will cheat you blind. It's their 'culture' (post cultural revolution), don't blame the poor Chinese pee-pull, not their fault.

Thor

 

[valve_integrator]

It's pretty obvious they measure parameters and likely match (pass/fail) drivers after running them for 2 hours. Sounds like good practice to me.

if you read their blurb they actually STATE they don't do this whatever "burn in" for goods that go out the door...
There you hit the good old "pot luck", and blurb bs cycle.

Btw my RA drivers are more prone after decades to "burn out" rather than "burn in", however there I sometimes wonder how with 2 bass RF resonant ports for the 12" and 15" drivers in a nr 250L speaker cab, whether my hunch was correct.

Lots of the best solutions work by hunches, inc holding my fingers in my ears when playing in a PRC concert hall..
..lots of stuff being supported these days by "bling" and sheer BS (eg, the PRC oriental arts centre, which they cracked out to b one of the best sounding in the world like that ultra crap Mariinski 2, also mega-hyped.)

If they can't make a concert hall sound any good, F-A chance of getting a speaker that works anything other than by marketing b..llox! huh!

 

[valve_integrator]

I have been at the mixing desk for concerts in china, with all Chinese made gear.

I was able to obtain excellent sound quality, no acoustic feedback, monitors and backline excellent clarity and way loud enough according to the musicians, excellent sound in the house.

The hardware is generally pretty decent, it's the wet ware that is at fault and reason for the usually absolutely terrible results.

yes it's the engineers were always crap, because they hadn't the slightest idea what sounded good.
I remember wrestling with the guys to try to force a given solution up to 2mins before I was due on stage...(or using my stuff A-B or X-Y style hidden away in a flower pot with my notebook stowed doing DTD over XLR.)

In this case it was the latest and greatest western equipment....I was always astonished to see them multi miking us with B & K or DPA mega expenses stuff, but trying to stick 30 of them on stage so they practically recorded the hairs in my nose rubbing together....

Funnily enough the video team knew what they were doing....must come from doing long times filming Xi and co at party confs.

 

[Thorsten Loesch]

Does this break-in (would not call it burn-in which is a different thing in my mind) that minutes, hours, days, weeks, months, years or decades ?

It typically is the "power limit test" signal according to whatever specific standard (there are many, you can tell them whichever to use for your product) for several hours, usually 4-8 prior to testing T/S parameters and FR.

Differences depend upon specific design of drivers etc. You can also ask them to give you an extra set of "green" measurements to use for incoming QC, so there is no confusion.As the factories usually do not "break in / burn in / whatever-in" production drivers.

What would be the average time before one can use it. (The linked data sheet states 2 hours and 10 hours 'rest')
That would be a far cry from the years some make it out to be.

10V RMS with 20Hz creates a lot of movement. Hence it used.

If listening with music at normal levels there is much less "exercise".

My personal regime of preconditioning emitting transducers uses 20Hz-20kHz bandwidth limited mono pink noise at levels that drive the woofer to excursion limits but do not exceed rated power limits.

Wire speakers in opposite polarity and place face to face to limit noise, "cook" for 48-72 hours.

A question would be would headphone drivers be affected in a similar way ? Very different construction, materials, scale, mass, 'beating' it has gotten.

The same principle effects apply. I also "precondition" headphones. Some change a lot, least change is observed in 'stats, e-stats show no change, m-stats are at confidence levels for my test gear but dynamic headphones often change a fair bit.

I have measured new headphones, for instance, right out of the box (without listening to it first), played with them, measured them again and even measured them years later again (same fixture, same everything) and found no real differences that could not be explained by differences in positioning, pad wear.

Well, as said, it depends a lot. Some dynamic headphone drivers are more like miniature speakers, others are more like microphones in reverse.

Speakers... I am sure there is break-in. But how long does it take is the biggest question. I mean, how can one tune a bass port if the speaker resonance point keeps on changing?

Well, there is that. If you tune with a "green" driver and maximally flat, your bass response "run in" may be up a dB or two.

Other than anecdotal (mine is obvious also anecdotal) do you have any official documentation you can share, say a now obsolete product, which would not be under an NDA ?

Documentation for what?

Many Factory customer approval sheets float around on AliExpress etc. for speskers. Just have a read and read the small print.

As said, Wavecor are very upfront about this. You can always compare Vance Dickinson's measurements of the same drivers in voice coil.

Also, iirc, he too has a "preconditioning" regime before measuring drivers, probably less time consuming than my own, as he measures drivers for a living, while I do it for engineering and verification purposes.

Thor

 

[Thorsten Loesch]

Lots of the best solutions work by hunches, inc holding my fingers in my ears when playing in a PRC concert hall..

I prefer well designed ear plugs with 37dB attenuation and a fairly flat frequency response.

If they can't make a concert hall sound any good, F-A chance of getting a speaker that works anything other than by marketing b..llox! huh!

Well, they can assemble generic parts (that everyone else uses as well) into drivers just as well as any other place and in the PRC nobody cares if the glue used in the driver's or voice coils gives the workers cancer, labour turnover is very anyway, so long term exposure for many years is rare.

As Dorothy said, when she first set foot into the middle kingdom:

Thor

 

[kongwee]

If you do retail, you won't sell off your demo set.

 

[solderdude]

My personal regime of preconditioning emitting transducers uses 20Hz-20kHz bandwidth limited mono pink noise at levels that drive the woofer to excursion limits but do not exceed rated power limits.

Wire speakers in opposite polarity and place face to face to limit noise, "cook" for 48-72 hours.

I like over-kill and over dimensioning when there is no info available. But what does research show.
I meant manufacturers will certainly have done durability and aging testing revealing such info.
Do you happen to own a copy of such research or can share some numbers of such articles ?

 

[egellings]

Actually, "burn in" that is change of characteristics with time and use towards design center parsmeters, including operational ones, are well documented for almost all electronic, electric and electromechanical (and mechanical) components.

So I really cannot understand the debate. Burn in is an established fact. Deal with it. If that puts your world view into question, maybe your world view required correction (and rectification)?

V.I.T.R.I.O.L. - "Visita Interiora Terræ Rectificando Invenies Occultum Lapidem"

So if there is any tribalism and self deception it obviously goes like this: "break-in must NEVER EVER work, all my friends say so, I've done it, no harm no foul, and now I'm part of a special club!"

How audible such changes may be can be debated. Their existence cannot.

But there is little agreement on what is reliably audible and what is reliably inaudible and what are edge cases, and what scientific research is admissiable the to the debate that any debate resembles that of a cargo-cultitist berating a witchdoctor on a pacific island for being unscientific.

Thor

Please translate the Latin sentence. It's been decades since I took it in high school.

 

[Andretti60]

You may want to read the thread a bit more careful if that's what you got from it. -Unless you are just here for cheap internet points from people sharing your opinion and making fun, then carry on your wayward, son.

I asked a very simple question, you did not answer (do you have an answer) and decided to patronize me.
Please do not call me “son”.
PS what does “cheap internet points” mean.

 

[Thorsten Loesch]

Please translate the Latin sentence. It's been decades since I took it in high school.

"V.I.T.R.I.O.L." is an acronym for "Visita Interiora Terræ Rectificando Invenies Occultum Lapidem", an alchemic motto that can be roughly translated as "Visit the interior of the earth and [by] rectifying [yourself] you will find the hidden stone".

Thor

PS, I'm not an Alchemist, just well read.

 

[Deleted member 48726]

I asked a very simple question, you did not answer (do you have an answer) and decided to patronize me.
Please do not call me “son”.
PS what does “cheap internet points” mean.

It wasn't really an honest question, now was it? Don't act all surprised to get a cheeky answer to that post.

If “burn-in” changes the parameters of a component, how come does it always do for the better? I never heard anyone complaining that after 201 hours, 32 minutes and 52 seconds of burn-in a speaker was sounding bad.

 

[Andretti60]

It wasn't really an honest question, now was it? Don't act all surprised to get a cheeky answer to that post.

And why do you think it was not a “honest” question.
You claim that speakers sound changes after a period of time, so I repeat the question: why always for the better.